WO2019011188A1 - Method and device for processing measured data of inertial measurement unit, and unmanned aerial vehicle - Google Patents

Method and device for processing measured data of inertial measurement unit, and unmanned aerial vehicle Download PDF

Info

Publication number
WO2019011188A1
WO2019011188A1 PCT/CN2018/094805 CN2018094805W WO2019011188A1 WO 2019011188 A1 WO2019011188 A1 WO 2019011188A1 CN 2018094805 W CN2018094805 W CN 2018094805W WO 2019011188 A1 WO2019011188 A1 WO 2019011188A1
Authority
WO
WIPO (PCT)
Prior art keywords
model
error
inertial measurement
measurement unit
data
Prior art date
Application number
PCT/CN2018/094805
Other languages
French (fr)
Chinese (zh)
Inventor
于斌
Original Assignee
深圳市道通智能航空技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市道通智能航空技术有限公司 filed Critical 深圳市道通智能航空技术有限公司
Publication of WO2019011188A1 publication Critical patent/WO2019011188A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
    • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
    • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
    • G01C21/183—Compensation of inertial measurements, e.g. for temperature effects
    • G01C21/188—Compensation of inertial measurements, e.g. for temperature effects for accumulated errors, e.g. by coupling inertial systems with absolute positioning systems
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
    • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
    • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
    • G01C21/18—Stabilised platforms, e.g. by gyroscope

Definitions

  • the present invention relates to the field of communications, and in particular, to a method, device, and drone for measuring data processing of an inertial measurement unit.
  • MEMS micro-motor system
  • IMU inertial measurement unit
  • a method for processing measurement data of an inertial measurement unit includes the following steps:
  • the measurement data is filtered by the filtering model to reduce an error of the measurement data.
  • the method prior to the establishing the error model, further comprises establishing an Allan variance equation using the measured data of the inertial measurement unit, the Alon variance equation being formulated as:
  • the error ⁇ ( ⁇ ) of the Allen's variance equation satisfies the following formula:
  • the error model first-order Gauss Markov model + angular rate walk-around noise model + white noise model;
  • the first-order Gauss Markov model is expressed as:
  • the angular rate walkaway noise model is expressed as:
  • the white noise model is expressed as: ⁇ WN
  • the filtering model is formulated as:
  • x acc is the state quantity of the acceleration
  • x gyro is the state quantity of the gyroscope. Obtained according to the error model
  • the filtering processing the measurement data according to the filtering model comprises:
  • k is the preset time, k+1 is the next time, X k is the value of x in equation (1) at time K, P o is the covariance matrix of the state, A is the system matrix, and A T is Transpose of the system matrix, Q is the covariance matrix of the process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimensions.
  • An apparatus for measuring data processing by an inertial measurement unit comprising:
  • An inertial measurement unit configured to acquire measurement data of the inertial measurement unit
  • a filtering model establishing unit configured to establish a filtering model according to the error model
  • a processing unit configured to filter the measurement data by using the filtering model to reduce an error of the measurement data.
  • the processing device further includes:
  • the Allen variance equation establishing unit is configured to establish an Allan variance equation by using the measured data of the inertial measurement unit, and the Allen variance equation established by the Allan variance equation establishing unit is formulated as:
  • the error ⁇ ( ⁇ ) of the Allen's variance equation satisfies the following formula:
  • error model first-order Gauss Markov model + angular rate walk-around noise model + white noise model
  • the first-order Gauss Markov model is expressed as:
  • the angular rate walkaway noise model is expressed as:
  • the white noise model is expressed as: ⁇ WN
  • the filtering model established by the filtering model establishing unit is formulated as:
  • x acc is the state quantity of the acceleration
  • x gyro is the state quantity of the gyroscope. Obtained according to the error model
  • the filtering processing the measurement data according to the filtering model comprises:
  • k is the preset time, k+1 is the next time, X k is the value of x in equation (1) at time K, P o is the covariance matrix of the state, A is the system matrix, and A T is Transpose of the system matrix, Q is the covariance matrix of the process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimensions.
  • a drone comprising the device and the main control chip of any of the above, the device being connected to the main control chip and configured to transmit the processed measurement data to the main control chip.
  • the method acquires measurement data according to the inertial measurement unit, and parameterizes the error source of the IMU sensor, and then processes the measurement data according to the error model and the filtering model.
  • the accuracy of the sensor is improved and the main noise of the MEMS sensor is eliminated, and no additional economic cost is added, and no new circuit is added, which is advantageous for improving the flight performance and other related performance of the drone.
  • 1 is a flow chart of a method for measuring data processing by an inertial measurement unit in an embodiment
  • FIG. 2 is a graph showing the x-axis raw data of the accelerometer and the fitted data in an embodiment
  • 3 is a graph of y-axis raw data and post-fit data of an accelerometer in an embodiment
  • Figure 5 is a block diagram of an apparatus for measuring data processing of an inertial measurement unit in an embodiment
  • Figure 6 is a block diagram of an unmanned aerial vehicle in an embodiment
  • FIG. 7 is a schematic diagram of raw data of three axes x, y, and z of an accelerometer in an embodiment
  • FIG. 8 is a schematic diagram of raw data of a three-axis x, y, and z of a gyroscope in an embodiment
  • FIG. 9 is a schematic diagram of filtered data of three axes x, y, and z of an accelerometer in an embodiment
  • FIG. 10 is a schematic diagram of filtered data of the gyroscope's three axes x, y, and z in an embodiment.
  • FIG. 1 is a flow chart of a method of processing measurement data of an inertial measurement unit in an embodiment, the method including steps S110-S150. among them:
  • K min the smallest number of packets in another trial.
  • nk+i is the subscript of the number of recorded data sets in the word test.
  • the data is used to calculate the difference above.
  • the maximum number of sampling points in a single group should not exceed N/2.
  • the error of the variance also increases. Therefore, the error of the Allan variance equation should satisfy the following formula:
  • the error should be kept in a certain area. For example, if the error is not more than 5%, the error data can be changed according to the actual situation, for example, it can be no more than 1%, 3%, 6%, 8% or 10% error. If you choose not to exceed 5% of the error, K should satisfy K>201. In this case, the maximum number of sampling points in a single data group satisfies n max ⁇ N/201. Maximum correlation time Where T is the total time of data collection.
  • the error model is formulated as:
  • Error model first-order Gauss Markov model + angular rate walk-off noise model + white noise model.
  • the first-order Gauss-Markov model is expressed as:
  • the angular velocity walk model is:
  • ⁇ t is the sampling time interval
  • k is the time
  • T c is the relevant time
  • T c is the relevant time
  • the Normal distribution Normal distribution And ⁇ WN conforms to a normal distribution
  • the fitting effect of the three axes x, y, and z of the accelerometer in one test is shown in FIG. 2 to FIG. 4 .
  • the X1 curve in FIG. 2 is a graph of the original data of the accelerometer x-axis
  • FIG. 2 The middle X2 curve is a graph of the x-axis fitting data of the accelerometer
  • the Y1 curve in Fig. 3 is a graph of the accelerometer x raw data
  • the Y2 curve in Fig. 3 is a graph of the accelerometer x fitted data
  • the middle Z1 curve is a graph of the z-axis raw data of the accelerometer
  • the Z2 curve in FIG. 4 is a graph of the z-axis fitted data of the accelerometer.
  • the other vertical axes in Figures 2 - 4 are standard allan Variance sigma, and the horizontal axis is Time cluster.
  • x acc is the state quantity of the acceleration, Obtained according to the error model, formula (1), ie Corresponding to the formula (1)
  • the subscript acc represents the relevant amount of the accelerometer.
  • x gyro the state quantity of the gyroscope, Obtained according to formula (1), ie Also corresponds to the formula (1)
  • the subscript gyro indicates the relevant amount of the gyroscope.
  • the actual measurement data is:
  • the actual output value of the filter is:
  • S150 Filtering the measurement data of the inertial measurement unit by using the filtering model to reduce the error of the measurement data, which may include the following steps:
  • the first step is to update the status according to the following formula:
  • the second step updates the measured values according to the following formula and completes the filtering:
  • the subscript k in the above formula represents time k, that is, k is a preset time. K+1 is the next moment, x k is the value of x at the kth moment in equation (1), P o is the covariance matrix of the state, A is the system matrix, A T is the transpose of the system matrix, Q is Covariance matrix of process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimension.
  • X1, Y1, and Z1 are schematic diagrams of raw data of acceleration X-axis, Y-axis, and Z-axis, as shown in Fig. 8.
  • X2, Y2, and Z3 are gyroscope X-axis and Y-axis. Schematic diagram of the raw data of the Z-axis. The error of the measurement data is reduced after the processing of this embodiment.
  • X3, Y3, and Z3 in the figure are schematic diagrams of processed data of acceleration X-axis, Y-axis, and Z-axis, as shown in FIG. 10, in which X4, Y4, and Z4 are gyroscopes X-axis and Y.
  • the IMU sensor error source is parameterized to eliminate sensor noise and long-term errors. It can improve the accuracy of the sensor, eliminate the main noise of the MEMS sensor, add no additional economic cost, and not add new circuits, reducing the cost of the product. It can improve the output accuracy of low-cost MEMS IMU, improve the flight performance and other related performance of the drone, and is easy to apply on a large scale.
  • the filtering model may employ a particle filtering processing model. That is to say, based on the filtering mode of the filtering model, other filtering methods such as particle filtering can be selected.
  • the sensor error model can be established in other ways, such as a first-order Markov fit.
  • FIG. 5 is a block diagram of a processing apparatus for measuring data of an inertial measurement unit in an embodiment, the apparatus including an inertial measurement unit 100, an Allen variance equation establishing unit 700, an error model establishing unit 200, a filtering model establishing unit 300, and a processing unit 400.
  • the inertial measurement unit 100 is configured to acquire measurement data of the inertial measurement unit.
  • the Alan variance equation building unit 700 is used to establish the Allen variance equation.
  • K min the smallest number of packets in another trial.
  • nk+i is the subscript of the number of recorded data sets in the word test.
  • the data is used to calculate the difference above.
  • the maximum number of sampling points in a single group should not exceed N/2.
  • the error of the variance also increases. Therefore, the error of the Allan variance equation should satisfy the following formula:
  • the error should be kept in a certain area. For example, if the error is not more than 5%, the error data can be changed according to the actual situation, for example, it can be no more than 1%, 3%, 6%, 8% or 10% error. If you choose not to exceed 5% of the error, K should satisfy K>201. In this case, the maximum number of sampling points in a single data group satisfies n max ⁇ N/201. Maximum correlation time Where T is the total time of data collection.
  • the error model establishing unit 200 is configured to establish an error model.
  • the error model is formulated as:
  • Error model first-order Gauss Markov model + angular rate walk-off noise model + white noise model.
  • the first-order Gauss-Markov model is expressed as:
  • the angular velocity walk model is:
  • ⁇ t is the sampling time interval
  • k is the time
  • T c is the relevant time
  • T c is the relevant time
  • the Normal distribution Normal distribution And ⁇ WN conforms to a normal distribution
  • the fitting effect of the three axes x, y, and z of the accelerometer in one test is shown in FIG. 2 to FIG. 4 .
  • the X1 curve in FIG. 2 is a graph of the original data of the accelerometer x-axis
  • FIG. 2 The middle X2 curve is a graph of the x-axis fitting data of the accelerometer
  • the Y1 curve in Fig. 3 is a graph of the accelerometer x raw data
  • the Y2 curve in Fig. 3 is a graph of the accelerometer x fitted data
  • the middle Z1 curve is a graph of the z-axis raw data of the accelerometer
  • the Z2 curve in FIG. 4 is a graph of the z-axis fitted data of the accelerometer.
  • the other vertical axes in Figures 2 - 4 are standard allan Variance sigma, and the horizontal axis is Time cluster.
  • the filtering model establishing unit 300 is configured to establish a filtering model according to the error model.
  • the filtering model is expressed as:
  • x acc is the state quantity of the acceleration, Obtained according to the error model, formula (1), ie Corresponding to the formula (1)
  • the subscript acc represents the relevant amount of the accelerometer.
  • x gyro the state quantity of the gyroscope, Obtained according to formula (1), ie Also corresponds to the formula (1)
  • the subscript gyro indicates the relevant amount of the gyroscope.
  • the actual measurement data is
  • the actual output value of the filter is:
  • the processing unit 400 is configured to perform filtering processing on the measurement data by using the filtering model to reduce an error of the measurement data. Specifically, the following steps may be included:
  • the first step is to update the status according to the following formula:
  • the second step updates the measured values according to the following formula and completes the filtering:
  • the subscript k in the above formula represents time k, that is, k is a preset time. K+1 is the next moment, x k is the value of x at the kth moment in equation (1), P o is the covariance matrix of the state, A is the system matrix, A T is the transpose of the system matrix, Q is Covariance matrix of process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimension.
  • X1, Y1, and Z1 are schematic diagrams of raw data of acceleration X-axis, Y-axis, and Z-axis, as shown in Fig. 8.
  • X2, Y2, and Z3 are gyroscope X-axis and Y-axis. Schematic diagram of the raw data of the Z-axis. The error of the measurement data is reduced after the processing of this embodiment.
  • X3, Y3, and Z3 in the figure are schematic diagrams of processed data of acceleration X-axis, Y-axis, and Z-axis, as shown in FIG. 10, in which X4, Y4, and Z4 are gyroscopes X-axis and Y.
  • FIG. 6 is a block diagram of an unmanned aerial vehicle including a processing device for measuring data of the inertial measurement unit and a main control chip 600.
  • the processing device for measuring data of the inertial measurement unit is connected to the main control chip 600, and is used for The processed measurement data is transmitted to the main control chip 600.
  • the UAV further includes a data fusion unit 500.
  • the data fusion unit 500 includes a data interface 510 and a sensor 520.
  • the data interface 510 is coupled to the processing unit 400.
  • the sensor may be a GPS, a visual sensor, a sonar, a radar, or the like.
  • the inertial measurement unit data processed by the processing unit 400 and the data acquired by other sensors are then data fused.
  • the fused data is then sent to the backend drone master chip 600. It is beneficial to improve the flight performance and other related performance of the drone.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Gyroscopes (AREA)

Abstract

A method and device for processing measured data of an inertial measurement unit (100), and an unmanned aerial vehicle. The method for processing measured data of an inertial measurement unit (100) comprises the following steps: obtaining measured data of an inertial measurement unit (100) (S110); establishing an error model (S130); establishing a filtering mode according to the error model (S140); and filtering the measured data by means of the filtering model to reduce the error of the measured data (S150). The measured data is obtained by means of the inertial measurement unit (100), and an error source of the inertial measurement unit (100) is parameterized; and then, the measured data is processed by means of the error model and the filtering model, to eliminate noise and long-term error of the inertial measurement unit (100). The method for processing measured data of an inertial measurement unit (100) can improve the output precision of the inertial measurement unit (100) of a low-cost micro-electromechanical system, and facilitate improving the flight performance of an unmanned aerial vehicle.

Description

惯性测量单元测量数据的处理方法、装置及无人机Method, device and drone for measuring measurement data of inertial measurement unit
相关申请的交叉引用Cross-reference to related applications
本申请要求申请号为201710558347.1，申请日为2017年7月10的中国专利申请的优先权，其全部内容通过引用结合于本文。The present application claims priority to Japanese Patent Application No. JP-A No. No. No. No. No. No.
技术领域Technical field
本发明涉及通信领域，特别是涉及一种惯性测量单元测量数据处理的方法、装置及无人机。The present invention relates to the field of communications, and in particular, to a method, device, and drone for measuring data processing of an inertial measurement unit.
背景技术Background technique
随着组合导航***的发展，低成本微电机***(MEMS)惯性测量单元(IMU)传感器得到了广泛的应用。低成本微电机***惯性测量单元具有成本低、体积小等优点。With the development of integrated navigation systems, low-cost micro-motor system (MEMS) inertial measurement unit (IMU) sensors have been widely used. Low-cost micro-motor system inertial measurement unit has the advantages of low cost and small size.
但是低成本的传感器产生的噪声比较大，而且随着时间的积累，传感器的误差也随着增大。However, the noise generated by low-cost sensors is relatively large, and the error of the sensor increases with time.
发明内容Summary of the invention
基于此，有必要提供一种误差小的惯性测量单元测量数据处理的方法、装置及无人机。Based on this, it is necessary to provide a method, a device and a drone for measuring the data processing of the inertial measurement unit with small error.
一种惯性测量单元测量数据的处理方法，包括以下步骤：A method for processing measurement data of an inertial measurement unit includes the following steps:
获取所述惯性测量单元的测量数据；Obtaining measurement data of the inertial measurement unit;
建立误差模型；Establish an error model;
根据所述误差模型建立滤波模型；Establishing a filtering model according to the error model;
通过所述滤波模型对所述测量数据进行滤波处理以减小所述测量数据的误差。The measurement data is filtered by the filtering model to reduce an error of the measurement data.
在其中一个实施例中，在所述建立误差模型之前，该方法还包括利用所述惯性测量单元的测量数据建立阿伦方差方程，所述阿伦方差方程用公式表示为：In one of the embodiments, prior to the establishing the error model, the method further comprises establishing an Allan variance equation using the measured data of the inertial measurement unit, the Alon variance equation being formulated as:
Figure PCTCN2018094805-appb-000001
Figure PCTCN2018094805-appb-000001
在其中一个实施例中，所述阿伦方差方程的误差σ(δ)满足以下公式：In one of the embodiments, the error σ(δ) of the Allen's variance equation satisfies the following formula:
Figure PCTCN2018094805-appb-000002
Figure PCTCN2018094805-appb-000002
在其中一个实施例中，所述误差模型＝一阶高斯马尔科夫模型+角速率游走噪声模型+白噪声模型；In one embodiment, the error model = first-order Gauss Markov model + angular rate walk-around noise model + white noise model;
所述一阶高斯马尔科夫模型表示为：The first-order Gauss Markov model is expressed as:
Figure PCTCN2018094805-appb-000003
Figure PCTCN2018094805-appb-000003
所述角速率游走噪声模型表示为：The angular rate walkaway noise model is expressed as:
Figure PCTCN2018094805-appb-000004
Figure PCTCN2018094805-appb-000004
所述白噪声模型表示为：ω WN The white noise model is expressed as: ω WN
因此，所述误差模型表示为：Therefore, the error model is expressed as:
Figure PCTCN2018094805-appb-000005
Figure PCTCN2018094805-appb-000005
其中，Δt为采样时间间隔，k为时刻，
Figure PCTCN2018094805-appb-000006
T c是相关时间，所述
Figure PCTCN2018094805-appb-000007
符合正态分布
Figure PCTCN2018094805-appb-000008
符合正态分布
Figure PCTCN2018094805-appb-000009
以及ω WN符合正态分布
Figure PCTCN2018094805-appb-000010
Where Δt is the sampling time interval and k is the time.
Figure PCTCN2018094805-appb-000006
T c is the relevant time, the
Figure PCTCN2018094805-appb-000007
Normal distribution
Figure PCTCN2018094805-appb-000008
Normal distribution
Figure PCTCN2018094805-appb-000009
And ω WN conforms to a normal distribution
Figure PCTCN2018094805-appb-000010
其中，
Figure PCTCN2018094805-appb-000011
以及ω WN通过对所述阿伦方差方程进行拟合获得。 among them,
Figure PCTCN2018094805-appb-000011
And ω WN is obtained by fitting the Allan variance equation.
在其中一个实施例中，所述滤波模型用公式表示为：In one of the embodiments, the filtering model is formulated as:
Figure PCTCN2018094805-appb-000012
Figure PCTCN2018094805-appb-000012
其中，x acc,为加速度的状态量，
Figure PCTCN2018094805-appb-000013
根据所述误差模型获取，x gyro,为陀螺仪的状态量，
Figure PCTCN2018094805-appb-000014
根据所述误差模型获取； Where x acc is the state quantity of the acceleration,
Figure PCTCN2018094805-appb-000013
According to the error model, x gyro is the state quantity of the gyroscope.
Figure PCTCN2018094805-appb-000014
Obtained according to the error model;
实际测量数据z＝[x acc x gyro]； Actual measurement data z=[x acc x gyro ];
则滤波的实际输出值为
Figure PCTCN2018094805-appb-000015
The actual output value of the filter
Figure PCTCN2018094805-appb-000015
在其中一个实施例中，所述根据所述滤波模型对测量数据进行滤波处理包括：In one embodiment, the filtering processing the measurement data according to the filtering model comprises:
根据以下公式进行状态更新：Status updates are based on the following formula:
Figure PCTCN2018094805-appb-000016
Figure PCTCN2018094805-appb-000016
P＝A*P 0*A T+Q； P=A*P 0 *A T +Q;
根据公式更新测量值并完成滤波：Update the measured value according to the formula and complete the filtering:
Figure PCTCN2018094805-appb-000017
Figure PCTCN2018094805-appb-000017
S＝H*P*H T+R S=H*P*H T +R
K＝P*H T/S K=P*H T /S
Figure PCTCN2018094805-appb-000018
Figure PCTCN2018094805-appb-000018
P 0＝(I 6-K*H)*P P 0 = (I 6 -K*H)*P
其中，k为预设时刻，k+1为下一时刻，X k为公式(1)中的x在第K时刻的值，P o为状态的协方差矩阵，A为***矩阵，A T为***矩阵的转置，Q为过程噪声的协方差矩阵，P为***状态变量的协方差矩阵，Z为测量值，H为输出矩阵，H T为输出矩阵的转置，R为噪声矩阵，I 6是单位矩阵且维度为6维。 Where k is the preset time, k+1 is the next time, X k is the value of x in equation (1) at time K, P o is the covariance matrix of the state, A is the system matrix, and A T is Transpose of the system matrix, Q is the covariance matrix of the process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimensions.
一种惯性测量单元测量数据处理的装置，包括：An apparatus for measuring data processing by an inertial measurement unit, comprising:
惯性测量单元，用于获取所述惯性测量单元的测量数据；An inertial measurement unit, configured to acquire measurement data of the inertial measurement unit;
误差模型建立单元，用于建立误差模型；An error model establishing unit for establishing an error model;
滤波模型建立单元，用于根据所述误差模型建立滤波模型；a filtering model establishing unit, configured to establish a filtering model according to the error model;
处理单元，用于通过所述滤波模型对所述测量数据进行滤波处理以减小所述测量数据的误差。And a processing unit, configured to filter the measurement data by using the filtering model to reduce an error of the measurement data.
在其中一个实施例中，该处理装置还包括：In one embodiment, the processing device further includes:
阿伦方差方程建立单元，用于利用所述惯性测量单元的测量数据建立阿伦方差方程，所述阿伦方差方程建立单元建立的所述阿伦方差方程用公式表示为：The Allen variance equation establishing unit is configured to establish an Allan variance equation by using the measured data of the inertial measurement unit, and the Allen variance equation established by the Allan variance equation establishing unit is formulated as:
Figure PCTCN2018094805-appb-000019
Figure PCTCN2018094805-appb-000019
在其中一个实施例中，所述阿伦方差方程的误差σ(δ)满足以下公式：In one of the embodiments, the error σ(δ) of the Allen's variance equation satisfies the following formula:
Figure PCTCN2018094805-appb-000020
Figure PCTCN2018094805-appb-000020
在其中一个实施例中，其中所述误差模型＝一阶高斯马尔科夫模型+角速率 游走噪声模型+白噪声模型；In one embodiment, wherein the error model = first-order Gauss Markov model + angular rate walk-around noise model + white noise model;
所述一阶高斯马尔科夫模型表示为：The first-order Gauss Markov model is expressed as:
Figure PCTCN2018094805-appb-000021
Figure PCTCN2018094805-appb-000021
所述角速率游走噪声模型表示为：The angular rate walkaway noise model is expressed as:
Figure PCTCN2018094805-appb-000022
Figure PCTCN2018094805-appb-000022
所述白噪声模型表示为：ω WN The white noise model is expressed as: ω WN
因此，所述误差模型表示为：Therefore, the error model is expressed as:
Figure PCTCN2018094805-appb-000023
Figure PCTCN2018094805-appb-000023
其中，Δt为采样时间间隔，k为时刻，
Figure PCTCN2018094805-appb-000024
T c是相关时间，所述
Figure PCTCN2018094805-appb-000025
符合正态分布
Figure PCTCN2018094805-appb-000026
符合正态分布
Figure PCTCN2018094805-appb-000027
以及ω WN符合正态分布
Figure PCTCN2018094805-appb-000028
Where Δt is the sampling time interval and k is the time.
Figure PCTCN2018094805-appb-000024
T c is the relevant time, the
Figure PCTCN2018094805-appb-000025
Normal distribution
Figure PCTCN2018094805-appb-000026
Normal distribution
Figure PCTCN2018094805-appb-000027
And ω WN conforms to a normal distribution
Figure PCTCN2018094805-appb-000028
其中，
Figure PCTCN2018094805-appb-000029
以及ω WN通过对所述阿伦方差方程进行拟合获得。 among them,
Figure PCTCN2018094805-appb-000029
And ω WN is obtained by fitting the Allan variance equation.
在其中一个实施例中，所述滤波模型建立单元建立的所述滤波模型用公式表示为：In one of the embodiments, the filtering model established by the filtering model establishing unit is formulated as:
Figure PCTCN2018094805-appb-000030
Figure PCTCN2018094805-appb-000030
其中，x acc,为加速度的状态量，
Figure PCTCN2018094805-appb-000031
根据所述误差模型获取，x gyro,为陀螺仪的状态量，
Figure PCTCN2018094805-appb-000032
根据所述误差模型获取； Where x acc is the state quantity of the acceleration,
Figure PCTCN2018094805-appb-000031
According to the error model, x gyro is the state quantity of the gyroscope.
Figure PCTCN2018094805-appb-000032
Obtained according to the error model;
实际测量数据z＝[x acc x gyro]； Actual measurement data z=[x acc x gyro ];
则滤波的实际输出值为
Figure PCTCN2018094805-appb-000033
The actual output value of the filter
Figure PCTCN2018094805-appb-000033
在其中一个实施例中，所述根据所述滤波模型对测量数据进行滤波处理包括：In one embodiment, the filtering processing the measurement data according to the filtering model comprises:
根据以下公式进行状态更新：Status updates are based on the following formula:
Figure PCTCN2018094805-appb-000034
Figure PCTCN2018094805-appb-000034
P＝A*P 0*A T+Q； P=A*P 0 *A T +Q;
根据公式更新测量值并完成滤波：Update the measured value according to the formula and complete the filtering:
Figure PCTCN2018094805-appb-000035
Figure PCTCN2018094805-appb-000035
S＝H*P*H T+R S=H*P*H T +R
K＝P*H T/S K=P*H T /S
Figure PCTCN2018094805-appb-000036
Figure PCTCN2018094805-appb-000036
P 0＝(I 6-K*H)*P P 0 = (I 6 -K*H)*P
其中，k为预设时刻，k+1为下一时刻，X k为公式(1)中的x在第K时刻的值，P o为状态的协方差矩阵，A为***矩阵，A T为***矩阵的转置，Q为过程噪声的协方差矩阵，P为***状态变量的协方差矩阵，Z为测量值，H为输出矩阵，H T为输出矩阵的转置，R为噪声矩阵，I 6是单位矩阵且维度为6维。 Where k is the preset time, k+1 is the next time, X k is the value of x in equation (1) at time K, P o is the covariance matrix of the state, A is the system matrix, and A T is Transpose of the system matrix, Q is the covariance matrix of the process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimensions.
一种无人机，包括上述任一所述的装置和主控芯片，所述装置与主控芯片连接，并用于将处理后的测量数据传送至主控芯片。A drone comprising the device and the main control chip of any of the above, the device being connected to the main control chip and configured to transmit the processed measurement data to the main control chip.
在上述惯性测量单元测量数据的处理方法、装置及无人机中，所述方法根据惯性测量单元获取测量数据，并将IMU传感器误差来源参数化，然后根据误差模型和滤波模型对测量数据进行处理，以消除传感器的噪声和长期误差。因此，提高了传感器的精度并消除了MEMS传感器的主要噪声，且不增加额外的经济成本，不添加新的电路，有利于提高无人机的飞行性能和其他相关性能。In the above method, device and UAV of the measurement data of the inertial measurement unit, the method acquires measurement data according to the inertial measurement unit, and parameterizes the error source of the IMU sensor, and then processes the measurement data according to the error model and the filtering model. To eliminate sensor noise and long-term errors. Therefore, the accuracy of the sensor is improved and the main noise of the MEMS sensor is eliminated, and no additional economic cost is added, and no new circuit is added, which is advantageous for improving the flight performance and other related performance of the drone.
附图说明DRAWINGS
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他实施例的附图。In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present invention, and those skilled in the art can obtain drawings of other embodiments according to the drawings without any creative work.
图1为一实施例中惯性测量单元测量数据处理的方法的流程图；1 is a flow chart of a method for measuring data processing by an inertial measurement unit in an embodiment;
图2为一实施例中加速度计x轴原始数据和拟合后数据的曲线图；2 is a graph showing the x-axis raw data of the accelerometer and the fitted data in an embodiment;
图3为一实施例中加速度计y轴原始数据和拟合后数据的曲线图；3 is a graph of y-axis raw data and post-fit data of an accelerometer in an embodiment;
图4为一实施例中加速度计z轴原始数据和拟合后数据的曲线图；4 is a graph of accelerometer z-axis raw data and post-fit data in an embodiment;
图5为一实施例中惯性测量单元测量数据处理的装置的框图；Figure 5 is a block diagram of an apparatus for measuring data processing of an inertial measurement unit in an embodiment;
图6为一实施例中无人机的框图；Figure 6 is a block diagram of an unmanned aerial vehicle in an embodiment;
图7为一实施例中加速度计三轴x、y、z的原始数据示意图；7 is a schematic diagram of raw data of three axes x, y, and z of an accelerometer in an embodiment;
图8为一实施例中陀螺仪三轴x、y、z的原始数据示意图；8 is a schematic diagram of raw data of a three-axis x, y, and z of a gyroscope in an embodiment;
图9为一实施例中加速度计三轴x、y、z的滤波后数据示意图；9 is a schematic diagram of filtered data of three axes x, y, and z of an accelerometer in an embodiment;
图10为一实施例中陀螺仪三轴x、y、z的滤波后数据示意图。FIG. 10 is a schematic diagram of filtered data of the gyroscope's three axes x, y, and z in an embodiment.
具体实施方式Detailed ways
为了便于理解本发明，下面将参照相关附图对本发明进行更全面的描述。附图中给出了本发明的较佳的实施例。但是，本发明可以以许多不同的形式来实现，并不限于本文所描述的实施例。相反地，提供这些实施例的目的是使对本发明的公开内容的理解更加透彻全面。In order to facilitate the understanding of the present invention, the present invention will be described more fully hereinafter with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention may be embodied in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the understanding of the present disclosure will be more fully understood.
需要说明的是，当元件被称为“固定于”另一个元件，它可以直接在另一个元件上或者也可以存在居中的元件。当一个元件被认为是“连接”另一个元件，它可以是直接连接到另一个元件或者可能同时存在居中元件。本文所使用的术语“垂直的”、“水平的”、“左”、“右”以及类似的表述只是为了说明的目的。It should be noted that when an element is referred to as being "fixed" to another element, it can be directly on the other element or the element can be present. When an element is considered to be "connected" to another element, it can be directly connected to the other element or. The terms "vertical," "horizontal," "left," "right," and the like, as used herein, are for illustrative purposes only.
除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。本文所使用的术语“及/或”包括一个或多个相关的所列项目的任意的和所有的组合。All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. The terminology used in the description of the present invention is for the purpose of describing particular embodiments and is not intended to limit the invention. The term "and/or" used herein includes any and all combinations of one or more of the associated listed items.
图1为一实施例中惯性测量单元测量数据的处理方法的流程图，该方法包括步骤S110-S150。其中：1 is a flow chart of a method of processing measurement data of an inertial measurement unit in an embodiment, the method including steps S110-S150. among them:
S110：获取惯性测量单元的测量数据；S110: Acquire measurement data of the inertial measurement unit;
S120、利用惯性测量单元的测量数据建立阿伦方差方程，该阿伦方差方程的建立过程如下：S120. Using the measurement data of the inertial measurement unit to establish an Allan variance equation, the process of establishing the Allen variance equation is as follows:
根据采样时间Ts、采集数据的时间长度T，获得总的采样点数N，N＝T/Ts。According to the sampling time Ts and the time length T of the collected data, the total number of sampling points N, N=T/Ts is obtained.
设n为在一次分组中的数据点的个数，n＝1,2,3,…,n max，
Figure PCTCN2018094805-appb-000037
Figure PCTCN2018094805-appb-000038
其中K min是再一次试验中最小的分组数。针对第k组数据,平均数表示为: Let n be the number of data points in a group, n=1, 2,3,...,n max ,
Figure PCTCN2018094805-appb-000037
Figure PCTCN2018094805-appb-000038
Where K min is the smallest number of packets in another trial. For the kth group of data, the average is expressed as:
Figure PCTCN2018094805-appb-000039
Figure PCTCN2018094805-appb-000039
其中nk+i是单词试验中记录数据组数的下标。Where nk+i is the subscript of the number of recorded data sets in the word test.
可以选择τ(n)＝n T s，则阿伦方差可以表示为： You can choose τ(n)=n T s , then the Allen variance can be expressed as:
Figure PCTCN2018094805-appb-000040
Figure PCTCN2018094805-appb-000040
从以上分析可以看到，K＝N/n依赖于单个分组中采样点的个数.因为K-1>0,最小值可以选择为K min＝2,这意味单次试验中最少有两组数据来进行以上方差的计算。单组中的最大采样点个数不应超过N/2.然而随着单组数据中采样点数的增加，方差的误差也变的增大，因此阿伦方差方程的误差应当满足以下公式: From the above analysis, it can be seen that K=N/n depends on the number of sampling points in a single group. Since K-1>0, the minimum value can be selected as K min =2, which means there are at least two groups in a single experiment. The data is used to calculate the difference above. The maximum number of sampling points in a single group should not exceed N/2. However, as the number of sampling points in a single set of data increases, the error of the variance also increases. Therefore, the error of the Allan variance equation should satisfy the following formula:
Figure PCTCN2018094805-appb-000041
Figure PCTCN2018094805-appb-000041
为了保证一定的精度，误差应该保持在一定的区域，例如，如果采用的是保持不超过5％的误差，当然这个误差数据可以根据实际情况改变的，如可以为不超过1％、3％、6％、8％或10％的误差。如果选择不超过5％的误差，K应该满足K>201。在这种情况下，单个数据组中的最大采样点数满足n max＜N/201。最大相关时间
Figure PCTCN2018094805-appb-000042
其中T是数据采集的总时间。 In order to ensure a certain precision, the error should be kept in a certain area. For example, if the error is not more than 5%, the error data can be changed according to the actual situation, for example, it can be no more than 1%, 3%, 6%, 8% or 10% error. If you choose not to exceed 5% of the error, K should satisfy K>201. In this case, the maximum number of sampling points in a single data group satisfies n max <N/201. Maximum correlation time
Figure PCTCN2018094805-appb-000042
Where T is the total time of data collection.
S130：建立误差模型，在其中一个实施例中，误差模型用公式表示为：S130: Establish an error model. In one embodiment, the error model is formulated as:
误差模型＝一阶高斯马尔科夫模型+角速率游走噪声模型+白噪声模型。Error model = first-order Gauss Markov model + angular rate walk-off noise model + white noise model.
其中，一阶高斯-马尔科夫模型表示为：Among them, the first-order Gauss-Markov model is expressed as:
Figure PCTCN2018094805-appb-000043
Figure PCTCN2018094805-appb-000043
角速度游走模型为：The angular velocity walk model is:
Figure PCTCN2018094805-appb-000044
Figure PCTCN2018094805-appb-000044
白噪声为ω WN； White noise is ω WN ;
因此，误差模型表示为：
Figure PCTCN2018094805-appb-000045
Therefore, the error model is expressed as:
Figure PCTCN2018094805-appb-000045
其中，Δt为采样时间间隔、k为时刻、
Figure PCTCN2018094805-appb-000046
T c是相关时间， Where Δt is the sampling time interval, k is the time,
Figure PCTCN2018094805-appb-000046
T c is the relevant time,
Figure PCTCN2018094805-appb-000047
T c是相关时间，所述
Figure PCTCN2018094805-appb-000048
符合正态分布
Figure PCTCN2018094805-appb-000049
符合正态分布
Figure PCTCN2018094805-appb-000050
以及ω WN符合正态分布
Figure PCTCN2018094805-appb-000051
Figure PCTCN2018094805-appb-000047
T c is the relevant time, the
Figure PCTCN2018094805-appb-000048
Normal distribution
Figure PCTCN2018094805-appb-000049
Normal distribution
Figure PCTCN2018094805-appb-000050
And ω WN conforms to a normal distribution
Figure PCTCN2018094805-appb-000051
其中，
Figure PCTCN2018094805-appb-000052
以及ω WN通过对所述阿伦方差方程进行拟合获得。 among them,
Figure PCTCN2018094805-appb-000052
And ω WN is obtained by fitting the Allan variance equation.
例如，加速度计三个轴x、y、z在一次试验中的拟合效果如图2至图4所示，具体的，图2中X1曲线为加速度计x轴原始数据的曲线图，图2中X2曲线为加速度计x轴拟合后数据的曲线图；图3中Y1曲线为加速度计x原始数据的曲线图，图3中Y2曲线为加速度计x拟合后数据的曲线图；图4中Z1曲线为加速度计z轴原始数据的曲线图，图4中Z2曲线为加速度计z轴拟合后数据的曲线图。其他图2-图4中的纵轴为标准阿伦方差西格玛(Standard allan Variance sigma)，横轴为时间群集(Time cluster)。For example, the fitting effect of the three axes x, y, and z of the accelerometer in one test is shown in FIG. 2 to FIG. 4 . Specifically, the X1 curve in FIG. 2 is a graph of the original data of the accelerometer x-axis, FIG. 2 The middle X2 curve is a graph of the x-axis fitting data of the accelerometer; the Y1 curve in Fig. 3 is a graph of the accelerometer x raw data, and the Y2 curve in Fig. 3 is a graph of the accelerometer x fitted data; The middle Z1 curve is a graph of the z-axis raw data of the accelerometer, and the Z2 curve in FIG. 4 is a graph of the z-axis fitted data of the accelerometer. The other vertical axes in Figures 2 - 4 are standard allan Variance sigma, and the horizontal axis is Time cluster.
S140：建立滤波模型，滤波模型用公式表示为：S140: Establish a filtering model, and the filtering model is expressed by a formula as:
Figure PCTCN2018094805-appb-000053
Figure PCTCN2018094805-appb-000053
其中，x acc,为加速度的状态量，
Figure PCTCN2018094805-appb-000054
根据误差模型即公式(1)获取，即
Figure PCTCN2018094805-appb-000055
相对应于公式(1)中的
Figure PCTCN2018094805-appb-000056
下标acc表示加速度计的有关量。x gyro，为陀螺仪的状态量，
Figure PCTCN2018094805-appb-000057
根据公式(1)获取，即
Figure PCTCN2018094805-appb-000058
也相对应于公式(1)中的
Figure PCTCN2018094805-appb-000059
下标gyro表示陀螺仪的有关量。 Where x acc is the state quantity of the acceleration,
Figure PCTCN2018094805-appb-000054
Obtained according to the error model, formula (1), ie
Figure PCTCN2018094805-appb-000055
Corresponding to the formula (1)
Figure PCTCN2018094805-appb-000056
The subscript acc represents the relevant amount of the accelerometer. x gyro , the state quantity of the gyroscope,
Figure PCTCN2018094805-appb-000057
Obtained according to formula (1), ie
Figure PCTCN2018094805-appb-000058
Also corresponds to the formula (1)
Figure PCTCN2018094805-appb-000059
The subscript gyro indicates the relevant amount of the gyroscope.
其中，实际测量数据为：Among them, the actual measurement data is:
z＝[x acc x gyro]； z=[x acc x gyro ];
滤波的实际输出值为：The actual output value of the filter is:
Figure PCTCN2018094805-appb-000060
Figure PCTCN2018094805-appb-000060
S150：通过上述滤波模型对惯性测量单元的测量数据进行滤波处理从而减小测 量数据的误差，可包括以下步骤：S150: Filtering the measurement data of the inertial measurement unit by using the filtering model to reduce the error of the measurement data, which may include the following steps:
第一步根据以下公式进行状态更新：The first step is to update the status according to the following formula:
Figure PCTCN2018094805-appb-000061
Figure PCTCN2018094805-appb-000061
P＝A*P 0*A T+Q； P=A*P 0 *A T +Q;
第二步根据以下公式更新测量值并完成滤波：The second step updates the measured values according to the following formula and completes the filtering:
Figure PCTCN2018094805-appb-000062
Figure PCTCN2018094805-appb-000062
S＝H*P*H T+R S=H*P*H T +R
K＝P*H T/S K=P*H T /S
Figure PCTCN2018094805-appb-000063
Figure PCTCN2018094805-appb-000063
P 0＝(I 6-K*H)*P P 0 = (I 6 -K*H)*P
其中，以上式子中的下标k，代表时刻k，即k为预设时刻。k+1为下一时刻，x k为公式(1)中的x在第k时刻的值，P o为状态的协方差矩阵，A为***矩阵，A T为***矩阵的转置，Q为过程噪声的协方差矩阵，P为***状态变量的协方差矩阵，Z为测量值，H为输出矩阵，H T为输出矩阵的转置，R为噪声矩阵，I 6是单位矩阵且维度为6维。 Wherein, the subscript k in the above formula represents time k, that is, k is a preset time. K+1 is the next moment, x k is the value of x at the kth moment in equation (1), P o is the covariance matrix of the state, A is the system matrix, A T is the transpose of the system matrix, Q is Covariance matrix of process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimension.
如图7所示，图中X1、Y1、Z1为加速度X轴、Y轴、Z轴的原始数据的示意图，如图8所示，图中X2、Y2、Z3为陀螺仪X轴、Y轴、Z轴的原始数据的示意图。经过本实施例的处理后减小所述测量数据的误差。如图9所示，图中X3、Y3、Z3为加速度X轴、Y轴、Z轴的处理后数据的示意图，如图10所示，图中X4、Y4、Z4为陀螺仪X轴、Y轴、Z轴的处理后数据的示意图。根据上述实施方式，可以将IMU的数据准度显著提高。将IMU传感器误差来源参数化，消除传感器的噪声和长期误差。可以提高传感器精度，消除MEMS传感器的主要噪声，不增加额外的经济成本，不添加新的电路，降低产品的成本。能够提高低成本MEMS IMU的输出准度，有利于提高无人机的飞行性能和其他相关性能，易于大规模应用。As shown in Fig. 7, X1, Y1, and Z1 are schematic diagrams of raw data of acceleration X-axis, Y-axis, and Z-axis, as shown in Fig. 8. In the figure, X2, Y2, and Z3 are gyroscope X-axis and Y-axis. Schematic diagram of the raw data of the Z-axis. The error of the measurement data is reduced after the processing of this embodiment. As shown in FIG. 9, X3, Y3, and Z3 in the figure are schematic diagrams of processed data of acceleration X-axis, Y-axis, and Z-axis, as shown in FIG. 10, in which X4, Y4, and Z4 are gyroscopes X-axis and Y. Schematic diagram of processed data for the axis and Z axis. According to the above embodiment, the data accuracy of the IMU can be significantly improved. The IMU sensor error source is parameterized to eliminate sensor noise and long-term errors. It can improve the accuracy of the sensor, eliminate the main noise of the MEMS sensor, add no additional economic cost, and not add new circuits, reducing the cost of the product. It can improve the output accuracy of low-cost MEMS IMU, improve the flight performance and other related performance of the drone, and is easy to apply on a large scale.
在一个实施例中，滤波模型可以采用粒子滤波处理模型。即基于滤波模型的滤波方式可以选择粒子滤波等其他滤波方式。In one embodiment, the filtering model may employ a particle filtering processing model. That is to say, based on the filtering mode of the filtering model, other filtering methods such as particle filtering can be selected.
在一个实施例中，传感器误差模型的建立，可以采用其他的方式，比如说一阶马尔科夫拟合的方式。In one embodiment, the sensor error model can be established in other ways, such as a first-order Markov fit.
图5为一实施例中惯性测量单元测量数据的处理装置的框图，该装置包括惯性测量单元100、阿伦方差方程建立单元700、误差模型建立单元200、滤波模型建立单元300和处理单元400。5 is a block diagram of a processing apparatus for measuring data of an inertial measurement unit in an embodiment, the apparatus including an inertial measurement unit 100, an Allen variance equation establishing unit 700, an error model establishing unit 200, a filtering model establishing unit 300, and a processing unit 400.
其中，惯性测量单元100，用于获取所述惯性测量单元的测量数据。The inertial measurement unit 100 is configured to acquire measurement data of the inertial measurement unit.
阿伦方差方程建立单元700：用于建立阿伦方差方程。获取阿伦方差包括：根据采样时间Ts、采集数据的时间长度T，获得总的采样点数N，N＝T/Ts。The Alan variance equation building unit 700: is used to establish the Allen variance equation. Obtaining the Allen variance includes: obtaining the total number of sampling points N, N=T/Ts according to the sampling time Ts and the time length T of the collected data.
设n为在一次分组中的数据点的个数，n＝1,2,3,…,n max，
Figure PCTCN2018094805-appb-000064
Figure PCTCN2018094805-appb-000065
其中K min是再一次试验中最小的分组数。针对第k组数据,平均数表示为: Let n be the number of data points in a group, n=1, 2,3,...,n max ,
Figure PCTCN2018094805-appb-000064
Figure PCTCN2018094805-appb-000065
Where K min is the smallest number of packets in another trial. For the kth group of data, the average is expressed as:
Figure PCTCN2018094805-appb-000066
Figure PCTCN2018094805-appb-000066
其中nk+i是单词试验中记录数据组数的下标。Where nk+i is the subscript of the number of recorded data sets in the word test.
可以选择τ(n)＝n T s，则阿伦方差可以表示为： You can choose τ(n)=n T s , then the Allen variance can be expressed as:
Figure PCTCN2018094805-appb-000067
Figure PCTCN2018094805-appb-000067
从以上分析可以看到，K＝N/n依赖于单个分组中采样点的个数.因为K-1>0,最小值可以选择为K min＝2,这意味单次试验中最少有两组数据来进行以上方差的计算。单组中的最大采样点个数不应超过N/2.然而随着单组数据中采样点数的增加，方差的误差也变的增大，因此阿伦方差方程的误差应当满足以下公式: From the above analysis, it can be seen that K=N/n depends on the number of sampling points in a single group. Since K-1>0, the minimum value can be selected as K min =2, which means there are at least two groups in a single experiment. The data is used to calculate the difference above. The maximum number of sampling points in a single group should not exceed N/2. However, as the number of sampling points in a single set of data increases, the error of the variance also increases. Therefore, the error of the Allan variance equation should satisfy the following formula:
Figure PCTCN2018094805-appb-000068
Figure PCTCN2018094805-appb-000068
为了保证一定的精度，误差应该保持在一定的区域，例如，如果采用的是保持不超过5％的误差，当然这个误差数据可以根据实际情况改变的，如可以为 不超过1％、3％、6％、8％或10％的误差。如果选择不超过5％的误差，K应该满足K>201。在这种情况下，单个数据组中的最大采样点数满足n max＜N/201。最大相关时间
Figure PCTCN2018094805-appb-000069
其中T是数据采集的总时间。 In order to ensure a certain precision, the error should be kept in a certain area. For example, if the error is not more than 5%, the error data can be changed according to the actual situation, for example, it can be no more than 1%, 3%, 6%, 8% or 10% error. If you choose not to exceed 5% of the error, K should satisfy K>201. In this case, the maximum number of sampling points in a single data group satisfies n max <N/201. Maximum correlation time
Figure PCTCN2018094805-appb-000069
Where T is the total time of data collection.
误差模型建立单元200，用于建立误差模型。在其中一个实施例中，误差模型用公式表示为：The error model establishing unit 200 is configured to establish an error model. In one of the embodiments, the error model is formulated as:
误差模型＝一阶高斯马尔科夫模型+角速率游走噪声模型+白噪声模型。Error model = first-order Gauss Markov model + angular rate walk-off noise model + white noise model.
其中，一阶高斯-马尔科夫模型表示为：Among them, the first-order Gauss-Markov model is expressed as:
Figure PCTCN2018094805-appb-000070
Figure PCTCN2018094805-appb-000070
角速度游走模型为：The angular velocity walk model is:
Figure PCTCN2018094805-appb-000071
Figure PCTCN2018094805-appb-000071
白噪声为ω WN； White noise is ω WN ;
因此，误差模型表示为：
Figure PCTCN2018094805-appb-000072
Therefore, the error model is expressed as:
Figure PCTCN2018094805-appb-000072
其中，Δt为采样时间间隔、k为时刻、
Figure PCTCN2018094805-appb-000073
T c是相关时间， Where Δt is the sampling time interval, k is the time,
Figure PCTCN2018094805-appb-000073
T c is the relevant time,
Figure PCTCN2018094805-appb-000074
T c是相关时间，所述
Figure PCTCN2018094805-appb-000075
符合正态分布
Figure PCTCN2018094805-appb-000076
符合正态分布
Figure PCTCN2018094805-appb-000077
以及ω WN符合正态分布
Figure PCTCN2018094805-appb-000078
Figure PCTCN2018094805-appb-000074
T c is the relevant time, the
Figure PCTCN2018094805-appb-000075
Normal distribution
Figure PCTCN2018094805-appb-000076
Normal distribution
Figure PCTCN2018094805-appb-000077
And ω WN conforms to a normal distribution
Figure PCTCN2018094805-appb-000078
其中，
Figure PCTCN2018094805-appb-000079
以及ω WN通过对所述阿伦方差方程进行拟合获得。 among them,
Figure PCTCN2018094805-appb-000079
And ω WN is obtained by fitting the Allan variance equation.
例如，加速度计三个轴x、y、z在一次试验中的拟合效果如图2至图4所示，具体的，图2中X1曲线为加速度计x轴原始数据的曲线图，图2中X2曲线为加速度计x轴拟合后数据的曲线图；图3中Y1曲线为加速度计x原始数据的曲线图，图3中Y2曲线为加速度计x拟合后数据的曲线图；图4中Z1曲线为加速度计z轴原始数据的曲线图，图4中Z2曲线为加速度计z轴拟合后数据的曲线图。其他图2-图4中的纵轴为标准阿伦方差西格玛(Standard allan Variance sigma)，横轴为时间群集(Time cluster)。For example, the fitting effect of the three axes x, y, and z of the accelerometer in one test is shown in FIG. 2 to FIG. 4 . Specifically, the X1 curve in FIG. 2 is a graph of the original data of the accelerometer x-axis, FIG. 2 The middle X2 curve is a graph of the x-axis fitting data of the accelerometer; the Y1 curve in Fig. 3 is a graph of the accelerometer x raw data, and the Y2 curve in Fig. 3 is a graph of the accelerometer x fitted data; The middle Z1 curve is a graph of the z-axis raw data of the accelerometer, and the Z2 curve in FIG. 4 is a graph of the z-axis fitted data of the accelerometer. The other vertical axes in Figures 2 - 4 are standard allan Variance sigma, and the horizontal axis is Time cluster.
滤波模型建立单元300，用于根据所述误差模型建立滤波模型。滤波模型用公式表示为：The filtering model establishing unit 300 is configured to establish a filtering model according to the error model. The filtering model is expressed as:
Figure PCTCN2018094805-appb-000080
Figure PCTCN2018094805-appb-000080
其中，x acc,为加速度的状态量，
Figure PCTCN2018094805-appb-000081
根据误差模型即公式(1)获取，即
Figure PCTCN2018094805-appb-000082
相对应于公式(1)中的
Figure PCTCN2018094805-appb-000083
下标acc表示加速度计的有关量。x gyro,为陀螺仪的状态量，
Figure PCTCN2018094805-appb-000084
根据公式(1)获取，即
Figure PCTCN2018094805-appb-000085
也相对应于公式(1)中的
Figure PCTCN2018094805-appb-000086
下标gyro表示陀螺仪的有关量。 Where x acc is the state quantity of the acceleration,
Figure PCTCN2018094805-appb-000081
Obtained according to the error model, formula (1), ie
Figure PCTCN2018094805-appb-000082
Corresponding to the formula (1)
Figure PCTCN2018094805-appb-000083
The subscript acc represents the relevant amount of the accelerometer. x gyro , the state quantity of the gyroscope,
Figure PCTCN2018094805-appb-000084
Obtained according to formula (1), ie
Figure PCTCN2018094805-appb-000085
Also corresponds to the formula (1)
Figure PCTCN2018094805-appb-000086
The subscript gyro indicates the relevant amount of the gyroscope.
其中，实际测量数据为Among them, the actual measurement data is
z＝[x acc x gyro]； z=[x acc x gyro ];
滤波的实际输出值为：The actual output value of the filter is:
Figure PCTCN2018094805-appb-000087
Figure PCTCN2018094805-appb-000087
处理单元400，用于通过所述滤波模型对所述测量数据进行滤波处理以减小所述测量数据的误差。具体可包括以下步骤：The processing unit 400 is configured to perform filtering processing on the measurement data by using the filtering model to reduce an error of the measurement data. Specifically, the following steps may be included:
第一步根据以下公式进行状态更新：The first step is to update the status according to the following formula:
Figure PCTCN2018094805-appb-000088
Figure PCTCN2018094805-appb-000088
P＝A*P 0*A T+Q； P=A*P 0 *A T +Q;
第二步根据以下公式更新测量值并完成滤波：The second step updates the measured values according to the following formula and completes the filtering:
Figure PCTCN2018094805-appb-000089
Figure PCTCN2018094805-appb-000089
S＝H*P*H T+R S=H*P*H T +R
K＝P*H T/S K=P*H T /S
Figure PCTCN2018094805-appb-000090
Figure PCTCN2018094805-appb-000090
P 0＝(I 6-K*H)*P P 0 = (I 6 -K*H)*P
其中，以上式子中的下标k，代表时刻k，即k为预设时刻。k+1为下一时刻，x k为公式(1)中的x在第k时刻的值，P o为状态的协方差矩阵，A为***矩阵，A T为***矩阵的转置，Q为过程噪声的协方差矩阵，P为***状态变量的协方 差矩阵，Z为测量值，H为输出矩阵，H T为输出矩阵的转置，R为噪声矩阵，I 6是单位矩阵且维度为6维。 Wherein, the subscript k in the above formula represents time k, that is, k is a preset time. K+1 is the next moment, x k is the value of x at the kth moment in equation (1), P o is the covariance matrix of the state, A is the system matrix, A T is the transpose of the system matrix, Q is Covariance matrix of process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimension.
如图7所示，图中X1、Y1、Z1为加速度X轴、Y轴、Z轴的原始数据的示意图，如图8所示，图中X2、Y2、Z3为陀螺仪X轴、Y轴、Z轴的原始数据的示意图。经过本实施例的处理后减小所述测量数据的误差。如图9所示，图中X3、Y3、Z3为加速度X轴、Y轴、Z轴的处理后数据的示意图，如图10所示，图中X4、Y4、Z4为陀螺仪X轴、Y轴、Z轴的处理后数据的示意图。根据上述实施方式，可以将IMU的数据准度显著提高。将IMU传感器误差来源参数化，然后根据误差模型和滤波模型对测量数据进行处理，消除传感器的噪声和长期误差。可以提高传感器精度，消除MEMS传感器的主要噪声，不增加额外的经济成本，不添加新的电路，降低产品的成本。能够提高低成本MEMS IMU的输出准度，有利于提高无人机的飞行性能和其他相关性能，易于大规模应用。图6为一实施例中无人机的框图，该无人机包括上述惯性测量单元测量数据的处理装置和主控芯片600，惯性测量单元测量数据的处理装置与主控芯片600连接，并用于将处理后的测量数据传送至主控芯片600。As shown in Fig. 7, X1, Y1, and Z1 are schematic diagrams of raw data of acceleration X-axis, Y-axis, and Z-axis, as shown in Fig. 8. In the figure, X2, Y2, and Z3 are gyroscope X-axis and Y-axis. Schematic diagram of the raw data of the Z-axis. The error of the measurement data is reduced after the processing of this embodiment. As shown in FIG. 9, X3, Y3, and Z3 in the figure are schematic diagrams of processed data of acceleration X-axis, Y-axis, and Z-axis, as shown in FIG. 10, in which X4, Y4, and Z4 are gyroscopes X-axis and Y. Schematic diagram of processed data for the axis and Z axis. According to the above embodiment, the data accuracy of the IMU can be significantly improved. The IMU sensor error source is parameterized, and then the measurement data is processed according to the error model and the filtering model to eliminate the noise and long-term error of the sensor. It can improve the accuracy of the sensor, eliminate the main noise of the MEMS sensor, add no additional economic cost, and not add new circuits, reducing the cost of the product. It can improve the output accuracy of low-cost MEMS IMU, improve the flight performance and other related performance of the drone, and is easy to apply on a large scale. 6 is a block diagram of an unmanned aerial vehicle including a processing device for measuring data of the inertial measurement unit and a main control chip 600. The processing device for measuring data of the inertial measurement unit is connected to the main control chip 600, and is used for The processed measurement data is transmitted to the main control chip 600.
该无人机还包括数据融合单元500，数据融合单元500包括数据接口510和传感器520，数据接口510连接处理单元400，传感器可以是GPS、视觉传感器、声呐、雷达等传感器。然后将处理单元400处理后的惯性测量单元数据和其他传感器获取的数据进行数据融合。然后将融合的数据发送给后端的无人机主控芯片600。有利于提高无人机的飞行性能和其他相关性能。The UAV further includes a data fusion unit 500. The data fusion unit 500 includes a data interface 510 and a sensor 520. The data interface 510 is coupled to the processing unit 400. The sensor may be a GPS, a visual sensor, a sonar, a radar, or the like. The inertial measurement unit data processed by the processing unit 400 and the data acquired by other sensors are then data fused. The fused data is then sent to the backend drone master chip 600. It is beneficial to improve the flight performance and other related performance of the drone.
以上所述实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。The technical features of the above-described embodiments may be arbitrarily combined. For the sake of brevity of description, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction between the combinations of these technical features, All should be considered as the scope of this manual.
以上所述实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权 利要求为准。The above-described embodiments are merely illustrative of several embodiments of the present invention, and the description thereof is more specific and detailed, but is not to be construed as limiting the scope of the invention. It should be noted that a number of variations and modifications may be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, the scope of protection of the invention should be subject to the appended claims.

Claims (13)

  1. 一种惯性测量单元测量数据的处理方法，其特征在于，包括以下步骤：A method for processing measurement data of an inertial measurement unit, comprising the steps of:
    获取所述惯性测量单元的测量数据；Obtaining measurement data of the inertial measurement unit;
    建立误差模型；Establish an error model;
    根据所述误差模型建立滤波模型；Establishing a filtering model according to the error model;
    通过所述滤波模型对所述测量数据进行滤波处理以减小所述测量数据的误差。The measurement data is filtered by the filtering model to reduce an error of the measurement data.
  2. 根据权利要求1所述的惯性测量单元测量数据的处理方法，其特征在于，在所述建立误差模型之前，该方法还包括：The method for processing measurement data of an inertial measurement unit according to claim 1, wherein before the establishing the error model, the method further comprises:
    利用所述惯性测量单元的测量数据建立阿伦方差方程，所述阿伦方差方程用公式表示为：An Allan variance equation is established using the measured data of the inertial measurement unit, which is formulated as:
    Figure PCTCN2018094805-appb-100001
    Figure PCTCN2018094805-appb-100001
  3. 根据权利要求2所述的惯性测量单元测量数据的处理方法，其特征在于，所述阿伦方差方程的误差σ(δ)满足以下公式：The method of processing inertial measurement unit measurement data according to claim 2, wherein the error σ(δ) of the Allan variance equation satisfies the following formula:
    Figure PCTCN2018094805-appb-100002
    Figure PCTCN2018094805-appb-100002
  4. 根据权利要求3所述的惯性测量单元测量数据的处理方法，其特征在于，所述误差模型＝一阶高斯马尔科夫模型+角速率游走噪声模型+白噪声模型；The method for processing measurement data of an inertial measurement unit according to claim 3, wherein the error model=first-order Gauss Markov model+angular rate walk-around noise model+white noise model;
    所述一阶高斯马尔科夫模型表示为：The first-order Gauss Markov model is expressed as:
    Figure PCTCN2018094805-appb-100003
    Figure PCTCN2018094805-appb-100003
    所述角速率游走噪声模型表示为：The angular rate walkaway noise model is expressed as:
    Figure PCTCN2018094805-appb-100004
    Figure PCTCN2018094805-appb-100004
    所述白噪声模型表示为：ω WN The white noise model is expressed as: ω WN
    因此，所述误差模型表示为：Therefore, the error model is expressed as:
    Figure PCTCN2018094805-appb-100005
    Figure PCTCN2018094805-appb-100005
    其中，Δt为采样时间间隔，k为时刻，
    Figure PCTCN2018094805-appb-100006
    T c是相关时间，所述
    Figure PCTCN2018094805-appb-100007
    符合正态分布
    Figure PCTCN2018094805-appb-100008
    符合正态分布
    Figure PCTCN2018094805-appb-100009
    以及ω WN符合正态分布
    Figure PCTCN2018094805-appb-100010
    Where Δt is the sampling time interval and k is the time.
    Figure PCTCN2018094805-appb-100006
    T c is the relevant time, the
    Figure PCTCN2018094805-appb-100007
    Normal distribution
    Figure PCTCN2018094805-appb-100008
    Normal distribution
    Figure PCTCN2018094805-appb-100009
    And ω WN conforms to a normal distribution
    Figure PCTCN2018094805-appb-100010
    其中，
    Figure PCTCN2018094805-appb-100011
    以及ω WN通过对所述阿伦方差方程进行拟合获得。     among them,
    Figure PCTCN2018094805-appb-100011
    And ω WN is obtained by fitting the Allan variance equation.
  5. 根据权利要求4所述的惯性测量单元测量数据的处理方法，其特征在于，所述滤波模型用公式表示为：The method for processing measurement data of an inertial measurement unit according to claim 4, wherein the filter model is expressed by:
    Figure PCTCN2018094805-appb-100012
    Figure PCTCN2018094805-appb-100012
    其中，x acc为加速度的状态量，
    Figure PCTCN2018094805-appb-100013
    根据所述误差模型获取，x gyro为陀螺仪的状态量，
    Figure PCTCN2018094805-appb-100014
    根据所述误差模型获取；     Where x acc is the state quantity of the acceleration,
    Figure PCTCN2018094805-appb-100013
    According to the error model, x gyro is the state quantity of the gyroscope,
    Figure PCTCN2018094805-appb-100014
    Obtained according to the error model;
    实际测量数据z＝[x acc x gyro]； Actual measurement data z=[x acc x gyro ];
    则滤波的实际输出值为
    Figure PCTCN2018094805-appb-100015
    The actual output value of the filter
    Figure PCTCN2018094805-appb-100015
  6. 根据权利要求5所述的惯性测量单元测量数据处理的方法，其特征在于，所述根据所述滤波模型对测量数据进行滤波处理包括：The method of processing data processing by the inertial measurement unit according to claim 5, wherein the filtering processing the measurement data according to the filtering model comprises:
    根据以下公式进行状态更新：Status updates are based on the following formula:
    Figure PCTCN2018094805-appb-100016
    Figure PCTCN2018094805-appb-100016
    P＝A*P 0*A T+Q； P=A*P 0 *A T +Q;
    根据公式更新测量值并完成滤波：Update the measured value according to the formula and complete the filtering:
    Figure PCTCN2018094805-appb-100017
    Figure PCTCN2018094805-appb-100017
    S＝H*P*H T+R S=H*P*H T +R
    K＝P*H T/S K=P*H T /S
    Figure PCTCN2018094805-appb-100018
    Figure PCTCN2018094805-appb-100018
    P 0＝(I 6-K*H)*P P 0 = (I 6 -K*H)*P
    其中，k为预设时刻，k+1为下一时刻，X k为公式(1)中的x在第K时刻的值，P o为状态的协方差矩阵，A为***矩阵，A T为***矩阵的转置，Q为过程噪声的协方差矩阵，P为***状态变量的协方差矩阵，Z为测量值，H为输出矩阵，H T为输出矩阵的转置，R为噪声矩阵，I 6是单位矩阵且维度为6维。 Where k is the preset time, k+1 is the next time, X k is the value of x in equation (1) at time K, P o is the covariance matrix of the state, A is the system matrix, and A T is Transpose of the system matrix, Q is the covariance matrix of the process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimensions.
  7. 一种惯性测量单元测量数据的处理装置，其特征在于，包括：A processing device for measuring data of an inertial measurement unit, comprising:
    惯性测量单元，用于获取所述惯性测量单元的测量数据；An inertial measurement unit, configured to acquire measurement data of the inertial measurement unit;
    误差模型建立单元，用于建立误差模型；An error model establishing unit for establishing an error model;
    滤波模型建立单元，用于根据所述误差模型建立滤波模型；a filtering model establishing unit, configured to establish a filtering model according to the error model;
    处理单元，用于通过所述滤波模型对所述测量数据进行滤波处理以减小所述测量数据的误差。And a processing unit, configured to filter the measurement data by using the filtering model to reduce an error of the measurement data.
  8. 根据权利要求7所述的惯性测量单元测量数据的处理装置，其特征在于，该处理装置还包括：The processing device for measuring the data of the inertial measurement unit according to claim 7, wherein the processing device further comprises:
    阿伦方差方程建立单元：用于利用所述惯性测量单元的测量数据建立阿伦方差方程，所述阿伦方差方程建立单元建立的所述阿伦方差方程用公式表示为：The Allen variance equation establishing unit is configured to establish an Allen variance equation by using the measured data of the inertial measurement unit, and the Allen variance equation established by the Allan variance equation establishing unit is formulated as:
    Figure PCTCN2018094805-appb-100019
    Figure PCTCN2018094805-appb-100019
  9. 根据权利要求8所述的惯性测量单元测量数据的处理装置，其特征在于，所述阿伦方差方程的误差σ(δ)满足以下公式：The apparatus for processing inertial measurement unit measurement data according to claim 8, wherein the error σ(δ) of the Allan variance equation satisfies the following formula:
    Figure PCTCN2018094805-appb-100020
    Figure PCTCN2018094805-appb-100020
  10. 根据权利要求9所述的惯性测量单元测量数据的处理装置，其特征在于，A processing apparatus for measuring data of an inertial measurement unit according to claim 9, wherein:
    所述误差模型建立单元建立的所述误差模型＝一阶高斯马尔科夫模型+角速率游走噪声模型+白噪声模型；The error model established by the error model establishing unit=first-order Gauss Markov model+angular rate walk-around noise model+white noise model;
    所述一阶高斯马尔科夫模型表示为：The first-order Gauss Markov model is expressed as:
    Figure PCTCN2018094805-appb-100021
    Figure PCTCN2018094805-appb-100021
    所述角速率游走噪声模型表示为：The angular rate walkaway noise model is expressed as:
    Figure PCTCN2018094805-appb-100022
    Figure PCTCN2018094805-appb-100022
    所述白噪声模型表示为：ω WN The white noise model is expressed as: ω WN
    因此，所述误差模型表示为：Therefore, the error model is expressed as:
    Figure PCTCN2018094805-appb-100023
    Figure PCTCN2018094805-appb-100023
    其中，Δt为采样时间间隔，k为时刻，
    Figure PCTCN2018094805-appb-100024
    T c是相关时间，所述
    Figure PCTCN2018094805-appb-100025
    符合正态分布
    Figure PCTCN2018094805-appb-100026
    符合正态分布
    Figure PCTCN2018094805-appb-100027
    以及ω WN符合正态分布
    Figure PCTCN2018094805-appb-100028
    Where Δt is the sampling time interval and k is the time.
    Figure PCTCN2018094805-appb-100024
    T c is the relevant time, the
    Figure PCTCN2018094805-appb-100025
    Normal distribution
    Figure PCTCN2018094805-appb-100026
    Normal distribution
    Figure PCTCN2018094805-appb-100027
    And ω WN conforms to a normal distribution
    Figure PCTCN2018094805-appb-100028
    其中，
    Figure PCTCN2018094805-appb-100029
    以及ω WN通过对所述阿伦方差方程进行拟合获得。     among them,
    Figure PCTCN2018094805-appb-100029
    And ω WN is obtained by fitting the Allan variance equation.
  11. 根据权利要求10所述的惯性测量单元测量数据的处理方法，其特征在于，所述滤波模型建立单元建立的所述滤波模型用公式表示为：The method for processing measurement data of an inertial measurement unit according to claim 10, wherein the filter model established by the filter model establishing unit is expressed by a formula:
    Figure PCTCN2018094805-appb-100030
    Figure PCTCN2018094805-appb-100030
    其中，x acc，为加速度的状态量，
    Figure PCTCN2018094805-appb-100031
    根据所述误差模型获取，x gyro，为陀螺仪的状态量，
    Figure PCTCN2018094805-appb-100032
    根据所述误差模型获取；     Where x acc is the state quantity of the acceleration,
    Figure PCTCN2018094805-appb-100031
    According to the error model, x gyro is the state quantity of the gyroscope.
    Figure PCTCN2018094805-appb-100032
    Obtained according to the error model;
    实际测量数据z＝[x acc x gyro]； Actual measurement data z=[x acc x gyro ];
    则滤波的实际输出值为
    Figure PCTCN2018094805-appb-100033
    The actual output value of the filter
    Figure PCTCN2018094805-appb-100033
  12. 根据权利要求11所述的惯性测量单元测量数据处理的方法，其特征在于，所述根据所述滤波模型对测量数据进行滤波处理包括：The method of processing data processing by the inertial measurement unit according to claim 11, wherein the filtering processing the measurement data according to the filtering model comprises:
    根据以下公式进行状态更新：Status updates are based on the following formula:
    Figure PCTCN2018094805-appb-100034
    Figure PCTCN2018094805-appb-100034
    P＝A*P 0*A T+Q； P=A*P 0 *A T +Q;
    根据公式更新测量值并完成滤波：Update the measured value according to the formula and complete the filtering:
    Figure PCTCN2018094805-appb-100035
    Figure PCTCN2018094805-appb-100035
    S＝H*P*H T+R S=H*P*H T +R
    K＝P*H T/S K=P*H T /S
    Figure PCTCN2018094805-appb-100036
    Figure PCTCN2018094805-appb-100036
    P 0＝(I 6-K*H)*P P 0 = (I 6 -K*H)*P
    其中，k为预设时刻，k+1为下一时刻，X k为公式(1)中的x在第K时刻的值，P o为状态的协方差矩阵，A为***矩阵，A T为***矩阵的转置，Q为过程噪声的协方差矩阵，P为***状态变量的协方差矩阵，Z为测量值，H为输出矩阵，H T为输出矩阵的转置，R为噪声矩阵，I 6是单位矩阵且维度为6维。 Where k is the preset time, k+1 is the next time, X k is the value of x in equation (1) at time K, P o is the covariance matrix of the state, A is the system matrix, and A T is Transpose of the system matrix, Q is the covariance matrix of the process noise, P is the covariance matrix of the system state variables, Z is the measured value, H is the output matrix, H T is the transpose of the output matrix, R is the noise matrix, I 6 is the identity matrix and the dimension is 6 dimensions.
  13. 一种无人机，其特征在于，包括如权利要求7-12任一所述的装置和主控芯片，所述装置与主控芯片连接，并用于将处理后的测量数据传送至主控芯片。An unmanned aerial vehicle, comprising: the device and the main control chip according to any one of claims 7-12, wherein the device is connected to the main control chip, and is configured to transmit the processed measurement data to the main control chip .
PCT/CN2018/094805 2017-07-10 2018-07-06 Method and device for processing measured data of inertial measurement unit, and unmanned aerial vehicle WO2019011188A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710558347.1 2017-07-10
CN201710558347.1A CN109238274B (en) 2017-07-10 2017-07-10 Processing method and device for measurement data of inertial measurement unit and unmanned aerial vehicle

Publications (1)

Publication Number Publication Date
WO2019011188A1 true WO2019011188A1 (en) 2019-01-17

Family

ID=65001877

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/094805 WO2019011188A1 (en) 2017-07-10 2018-07-06 Method and device for processing measured data of inertial measurement unit, and unmanned aerial vehicle

Country Status (2)

Country Link
CN (1) CN109238274B (en)
WO (1) WO2019011188A1 (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113091770B (en) * 2021-04-02 2023-12-15 上海有个机器人有限公司 Zero offset compensation method of inertial measurement sensor

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020008661A1 (en) * 2000-07-20 2002-01-24 Mccall Hiram Micro integrated global positioning system/inertial measurement unit system
CN101706284A (en) * 2009-11-09 2010-05-12 哈尔滨工程大学 Method for increasing position precision of optical fiber gyro strap-down inertial navigation system used by ship
CN101915579A (en) * 2010-07-15 2010-12-15 哈尔滨工程大学 Novel CKF(Crankshaft Fluctuation Sensor)-based SINS (Ship Inertial Navigation System) large misalignment angle initially-aligning method
CN102654406A (en) * 2012-04-11 2012-09-05 哈尔滨工程大学 Initial alignment method for moving bases based on combination of nonlinear prediction filtering and cubature Kalman filtering
CN104215262A (en) * 2014-08-29 2014-12-17 南京航空航天大学 On-line dynamic inertia sensor error identification method of inertia navigation system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020008661A1 (en) * 2000-07-20 2002-01-24 Mccall Hiram Micro integrated global positioning system/inertial measurement unit system
CN101706284A (en) * 2009-11-09 2010-05-12 哈尔滨工程大学 Method for increasing position precision of optical fiber gyro strap-down inertial navigation system used by ship
CN101915579A (en) * 2010-07-15 2010-12-15 哈尔滨工程大学 Novel CKF(Crankshaft Fluctuation Sensor)-based SINS (Ship Inertial Navigation System) large misalignment angle initially-aligning method
CN102654406A (en) * 2012-04-11 2012-09-05 哈尔滨工程大学 Initial alignment method for moving bases based on combination of nonlinear prediction filtering and cubature Kalman filtering
CN104215262A (en) * 2014-08-29 2014-12-17 南京航空航天大学 On-line dynamic inertia sensor error identification method of inertia navigation system

Also Published As

Publication number Publication date
CN109238274B (en) 2021-06-15
CN109238274A (en) 2019-01-18

Similar Documents

Publication Publication Date Title
WO2018076723A1 (en) Unmanned aerial vehicle flight control system based on cors network differential positioning
WO2018184467A1 (en) Method and device for detecting posture of ball head
Strohmeier et al. Ultra-wideband based pose estimation for small unmanned aerial vehicles
CN108981687B (en) Indoor positioning method with vision and inertia integration
CN108627152B (en) Navigation method of micro unmanned aerial vehicle based on multi-sensor data fusion
WO2021120525A1 (en) Navigation of unmanned device
CN107782304B (en) Mobile robot positioning method and device, mobile robot and storage medium
JP2021529947A (en) Motion sensor with drift correction
WO2018225067A1 (en) Fusion and calibration of sensor signals in a moving vehicle
CN106708088B (en) Coordinate calculation method and device, flight control method and system and unmanned aerial vehicle
CN108444468B (en) Directional compass integrating downward vision and inertial navigation information
CN112665574A (en) Underwater robot attitude acquisition method based on momentum gradient descent method
CN109521785A (en) It is a kind of to clap Smart Rotor aerocraft system with oneself
CN110595434B (en) Quaternion fusion attitude estimation method based on MEMS sensor
WO2019011188A1 (en) Method and device for processing measured data of inertial measurement unit, and unmanned aerial vehicle
CN113465596B (en) Four-rotor unmanned aerial vehicle positioning method based on multi-sensor fusion
CN113009816B (en) Method and device for determining time synchronization error, storage medium and electronic device
Oh et al. Extended Kalman filter with multi-frequency reference data for quadrotor navigation
CN113532481A (en) Zero offset compensation method of MEMS inertial navigation system
US20220306089A1 (en) Relative Position Tracking Using Motion Sensor With Drift Correction
CN103901459B (en) The filtering method of Measurement delay in a kind of MEMS/GPS integrated navigation system
Emran et al. A cascaded approach for quadrotor's attitude estimation
CN114413898B (en) Multi-sensor data fusion method and device, computer equipment and storage medium
TWI805141B (en) Positioning method and device for unmanned aerial vehicles
CN110892671A (en) Aircraft, data processing system and data processing method for aircraft

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18832493

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18832493

Country of ref document: EP

Kind code of ref document: A1